Amazon.com Inc. v. Hao
U.S. District Court, Western District of Washington, at Seattle · W.D. Wash. · Washington bar guidance
Conduct
A licensed attorney representing herself filed a motion to dismiss that the opposing party asserted cited nonexistent cases and misstated authorities; she acknowledged AI-assisted research.
Consequence
The court struck the motion with leave to refile on personally verified authorities, declined monetary sanctions at this time, and warned of future sanctions.
Lesson
Here the court held that a licensed attorney appearing pro se is held to Rule 11's objective standard, and that reliance on AI-assisted research is no excuse.
Verified September 30, 2026
- Citation
- Amazon.com Inc. v. Hao, No. 2:25-cv-02000-RSL, Order Resolving Amazon's Motion for an Order to Show Cause by Striking the Motion to Dismiss with Leave to Refile and Declining to Impose Monetary Sanctions at This Time (W.D. Wash. June 15, 2026), Dkt. 31
- Decided
- June 15, 2026
Summary
Amazon sued Kathy Q. Hao, "a U.S.-licensed attorney" according to its allegations, along with four Chinese companies and other defendants, alleging a scheme to obtain fraudulent trademark registrations and abuse Amazon's Brand Registry program. Hao, "proceeding pro se," moved to dismiss (Dkt. 6). Amazon opposed on the merits and separately moved for an order to show cause (Dkt. 13), identifying what it described as "fabricated, nonexistent, miscited, and materially misstated authorities" in the motion. The court summarized Amazon's assertion that the motion "cites nonexistent cases, pairs real reporter citations with unrelated or nonexistent case names, attributes propositions to cases that do not support them, and misstates statutory and regulatory authorities," and called the problems "serious." Hao filed a notice of errata (Dkt. 14) acknowledging that several cited authorities "were unreliable or incorrect," and in her opposition (Dkt. 17) acknowledged relying in part on AI-assisted research, stated that she believed the authorities were reliable, and pointed to her pro se status, her transactional and immigration practice, and her lack of federal litigation experience.
- AI tool:
- AI-assisted research, no tool named (the order records that Hao "relied in part on artificial-intelligence-assisted research, along with Westlaw and other research platforms")
- Sanction amount:
- No monetary sanction; the court struck the motion to dismiss without prejudice to refiling, declined "at this time" to impose monetary sanctions, and warned of future sanctions
What sanction did the court impose?
Judge Robert S. Lasnik held that Hao's pro se status did not excuse her failure to verify the authorities: pro se litigants are held to Rule 11's objective standard, and "Hao is not a layperson; she is a licensed attorney." Her asserted lack of federal litigation experience, financial stress, and "reliance on artificial-intelligence-assisted research do not excuse the filing of a motion containing unverified legal authorities." The court resolved Amazon's motion by striking the motion to dismiss without prejudice to refiling, directed that any renewed motion rely only on authorities Hao "has personally reviewed and verified," declined "at this time to impose monetary sanctions," and warned that any future filing containing fabricated, nonexistent, miscited, or materially misstated authority "may result in sanctions, including monetary sanctions, nonmonetary directives, striking of filings, or other relief the Court deems appropriate." A footnote explained that no separate order to show cause was entered because Hao's errata and opposition provided a sufficient response.
Why does Amazon.com Inc. v. Hao matter for law firms using AI?
The case arises from Amazon’s allegations that Kathy Q. Hao, a U.S.-licensed attorney, lent her name and USPTO credentials to foreign trademark-service providers in a scheme to obtain fraudulent trademark registrations. Hao represented herself and moved to dismiss. Amazon opposed on the merits and also moved for an order to show cause, cataloguing what it described as nonexistent cases, real reporter citations paired with the wrong case names, and misstated statutes and regulations in her motion.
The order is short and turns on one point: Hao’s status. She stressed that she was appearing pro se, practiced in transactional and immigration matters, and lacked federal litigation experience. The court answered that Rule 11’s standard is objective, that pro se litigants are held to it, and that in any event “Hao is not a layperson; she is a licensed attorney.” It rejected AI-assisted research as an excuse for filing unverified authorities.
The remedy was measured: the motion was struck with leave to refile on authorities Hao personally reviewed and verified, monetary sanctions were declined “at this time,” and the order warned of sanctions for any repeat. Firms documenting compliance may wish to consider that a lawyer’s own litigation, including matters where a lawyer is a party rather than counsel, may be held to the same verification standard as client work.
Implications for your firm
Operational steps a firm reading this case may wish to consider documenting. Strategic and rule-application calls belong to your firm's attorneys.
- Apply the firm's citation-verification practice to filings in matters where a lawyer is a party. Here the court held that a licensed attorney representing herself was not excused by her pro se status or her lack of federal litigation experience.
- Document verification of AI-assisted research against the cited sources before filing. The court held that reliance on AI-assisted research does not excuse filing unverified authorities, even alongside other research platforms.
- Consider that a court may act on an opposing party's motion without entering a separate show-cause order, relying on the filer's errata and opposition as the response.
Sources
Primary sources
- The order does not identify which citations were nonexistent or misstated; the description of the defects is the court's summary of Amazon's motion (Dkt. 13), which was not read. Hao's errata (Dkt. 14) and opposition (Dkt. 17) were not read; their content is reported as the order summarizes it.
- No AI tool is named. The AI reference is Hao's own acknowledgment as the order records it; the court made no separate finding about AI use.